refactor(search): address PR #750 round 11 review feedback

Three minor fixes from the round-11 review on PR #750:

- bm25_hybrid.py:209 — Comment said `doc_id` is `int (notes) or str (files)`,
  which is backwards. Notes, news_items, and deck_cards are stored as `str`
  (scanner.py:241, 666, 867); files are stored as `int` (scanner.py:425).
  Updated to point readers at scanner.py as the source of truth.

- verification.py:338 — Lowered the News-API 403/404 log line from `info`
  to `debug`. The News app being uninstalled or disabled is a predictable
  operational state (matching the other verifiers' debug-on-not-found
  paths), so this should not generate operator-dashboard noise. Transient
  errors immediately below stay at `warning` because they're unexpected.

- semantic.py:809 — `nc_get_vector_sync_status` was reading
  `document_receive_stream` via `getattr(..., None)`, but the attribute is
  guaranteed-defined on both `AppContext` and `OAuthAppContext` (as a
  field with `None` default). The defensive `getattr` masked typos that
  the eviction_task_group access at semantic.py:197-199 deliberately
  surfaces. Switched to direct access; the `if … is None:` value-check
  below is preserved (the attribute can legitimately be None before sync
  starts).

Items deliberately deferred (with rationale in the plan file):
- News verifier semaphore-hold during get_items (reviewer: "not required
  here, just worth tracking"; ADR already lists follow-ups).
- Hardcoded 2× over-fetch / VERIFICATION_OVERFETCH (TODO already in code).
- Integration test for the real Qdrant eviction filter (reviewer marked
  low-priority; type-preservation chain is unit-tested).

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

nextcloud_mcp_server/search/bm25_hybrid.py

@@ -208,7 +208,7 @@ class BM25HybridSearchAlgorithm(SearchAlgorithm):
             for result in search_response.points:
                 if result.payload is None:
                     continue
-                # doc_id can be int (notes) or str (files - file paths)
+                # doc_id can be int (files) or str (notes/news_items/deck_cards) — see scanner.py
                 doc_id = result.payload["doc_id"]
                 doc_type = result.payload.get("doc_type", "note")
                 chunk_start = result.payload.get("chunk_start_offset")

nextcloud_mcp_server/search/verification.py

@@ -335,7 +335,10 @@ async def _verify_news_items(
             # If the News API itself is gone (app disabled, user lost access),
             # treat *all* requested items as inaccessible. Eviction will reclaim.
             if _is_definitive_404_or_403(e):
-                logger.info(
+                # News app commonly disabled/uninstalled — debug-level keeps
+                # this off operator dashboards; transient errors below stay
+                # at warning because they're unexpected.
+                logger.debug(
                     "News API returned %s for user %s; treating all %d news_items as inaccessible",
                     e.response.status_code,
                     client.username,

nextcloud_mcp_server/server/semantic.py

@@ -804,11 +804,15 @@ def configure_semantic_tools(mcp: FastMCP):
             )
 
         try:
-            # Get document receive stream from lifespan context
+            # Get document receive stream from lifespan context. Direct
+            # attribute access matches the eviction_task_group pattern at
+            # ``nc_semantic_search`` (see comment there): both AppContext
+            # and OAuthAppContext define ``document_receive_stream``, so a
+            # missing attribute is a typo that should fail loudly. The
+            # value itself can legitimately be ``None`` before sync starts,
+            # which the check below handles.
             lifespan_ctx = ctx.request_context.lifespan_context
-            document_receive_stream = getattr(
-                lifespan_ctx, "document_receive_stream", None
-            )
+            document_receive_stream = lifespan_ctx.document_receive_stream
 
             if document_receive_stream is None:
                 logger.debug(